We used column chromatography, affinity binding, and bioassay methods to address whether the soluble tumor necrosis factor (TNF)-alpha receptors present in human colostrum and milk bind to and modify TNF-alpha bioactivity. In gel chromatography experiments, soluble TNF-alpha receptor I (sTNFRI) and sTNFRII in human colostrum sequentially increased their molecular sizes from 49 kD to 71 kD and 60 kD, respectively, after addition of increasing molar excesses of recombinant TNF-alpha. Application of colostrum to a TNF-alpha affinity matrix followed by washing and elution resulted in 2925-fold enrichment of sTNFRI, consistent with sTNFRI binding to the TNF-alpha affinity matrix. In other samples of colostrum and milk, the content of both sTNFRI and sTNFRII decreased significantly after passage over the matrix, but the material eluted from the matrix lost the ability to rebind to the TNF-alpha and was not active in a WEHI-13var bioassay for TNF-alpha. Specimens of human colostrum and milk diluted 1:16 shifted the LD50 for TNF-alpha 4-fold in this bioassay, and milk protection of WEHI-13var cells against TNF-alpha was significantly diminished after passage down the TNF-alpha affinity matrix (p < 0.001). Affinity purification of milk sTNFRI using polyclonal anti-sTNFRI produced fractions containing proteins of 30 kD, which could be visualized by Western blot using polyclonal anti-sTNFRI. Addition of this fraction to the WEHI-13var bioassay reversed the effects of 10 pg/mL TNF-alpha in the assay. These data demonstrate that sTNFRI and II from human colostrum and milk bind to TNF-alpha, that both colostrum and milk interfere with the bioactivity of TNF-alpha, and that affinity-purified sTNFRI from human milk blocks the bioactivity of TNF-alpha. These effects may contribute to the anti-inflammatory character of human colostrum and milk.